[x264-devel] Re: Rate control

[Måns Rullgård]

Loren Merritt <lorenm at u.washington.edu> writes:

> One more issue:
>
> Using the defaults (rc_sens=100, qp_step=4), bitrate completely fails to
> converge on some sources (I can send a log if needed).

You could also send some sources, or better yet send a patch.

> (Guess: The ratecontrol based on picture complexity is then stronger than
> the ratecontrol based on buffer fullness?)
> Setting either qp_step=1 or rc_sens < about 10 fixes this.
> (And the PSNR-optimal value of rc_sens was about 4.)

There is no theory behind the default rcsens value, I just picked
something that seemed to make some kind of sense and gave reasonable
results with my test cases.

-- 
Måns Rullgård
mru at mru.ath.cx

-- 
This is the x264-devel mailing-list
If you are in trouble, please contact <postmaster at videolan.org>